Base64 Decoder
Decode Base64 back into text. Invalid characters, bad length, and bytes that are not UTF-8 are reported here instead of being guessed.
About this tool
The decoder removes whitespace, checks that the remaining characters are in the standard Base64 alphabet, and adds = padding when the length is short by two or three characters. A remainder of one character is rejected. atob then turns the text into bytes, and TextDecoder reads those bytes as UTF-8 with fatal mode. Binary that is not valid UTF-8 fails with a clear message. This tool does not decode into a file download of arbitrary bytes.
URL-safe Base64 that uses - and _ is rejected. Convert those characters before decoding, or use standard Base64 from the encoder.
A successful decode is plain text, so text tools such as the Case Converter can follow it. The encoder is the inverse of this page, and the chain hides that inverse after a decode that was itself reached from the encoder. You can still open the encoder from the related links. A refresh can discard a handed-off value. It is not uploaded.
How to use it
- Paste Base64. Spaces and line breaks are ignored. Padding may be omitted when the length still makes sense.
- Decode it. A message explains malformed Base64 or bytes that are not UTF-8 text.
- Download the text or continue with a plain-text tool. The handoff stays in this browser.
Limitations
- Malformed Base64 is reported on the page. The decoder reads UTF-8 and does not guess another encoding.
Standard alphabet back to UTF-8 text
Whitespace is stripped, then the alphabet is checked
Standard Base64 is turned back into UTF-8 text. The paste is trimmed and internal whitespace is removed. The remaining characters have to be the standard alphabet: letters, digits, plus, slash, and equals signs used as padding. Minus and underscore, the URL-safe substitutes, are rejected. A length that leaves a remainder of one character after dividing by four is rejected. A remainder of two or three is padded with equals signs so the decoder can run. An empty box after trimming returns an empty string.
The bytes that come out of the Base64 decode are then interpreted as UTF-8, and the decode fails if those bytes are not valid UTF-8. A payload that was Base64 of an image, a zip, or any other non-text bytes will not become a string here. The error says the Base64 does not decode to UTF-8 text. That limit matches the encoder, which only ever writes UTF-8. This pair of pages is a text tool. It is not a general binary file converter.
Reversible spelling, not a lock
There is no key. Decoding does not prove that a trusted party produced the text. It only shows what the alphabet spelled. Do not treat a successful decode as authentication. A hash is not reversed here. If the box contains a JSON Web Token, the JWT page is the one that splits header and payload. This page will try to treat the whole paste as one Base64 blob and will fail when the dots and the URL-safe alphabet appear.
Padding mistakes
A missing equals sign at the end is added when the length allows it. An illegal character is not skipped. A line break in the middle is ignored as whitespace, which is why wrapped mail-style Base64 can work. A space that was actually supposed to be a plus must already be a plus: the strip removes whitespace, it does not turn spaces back into plus signs. If a transport changed plus signs into spaces, fix that before decoding. Use this page for standard Base64 that represents UTF-8 text.
Privacy
Your Base64 is processed in your browser and is not uploaded to UtilChain. The decoded text is not included in analytics.

